Two-stage object detection in low-light environments using deep learning image enhancement
This study presents a two-stage object detection system specifically tailored for low-light conditions. In the initial stage, supervised deep learning image enhancement techniques are utilized to improve image quality and enhance features. The second stage employs a computer vision algorithm for obj...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
PeerJ Inc.
2025-04-01
|
| Series: | PeerJ Computer Science |
| Subjects: | |
| Online Access: | https://peerj.com/articles/cs-2799.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1850217268174127104 |
|---|---|
| author | Ghaith Al-refai Hisham Elmoaqet Abdullah Al-Refai Ahmad Alzu’bi Tawfik Al-Hadhrami Abedalrhman Alkhateeb |
| author_facet | Ghaith Al-refai Hisham Elmoaqet Abdullah Al-Refai Ahmad Alzu’bi Tawfik Al-Hadhrami Abedalrhman Alkhateeb |
| author_sort | Ghaith Al-refai |
| collection | DOAJ |
| description | This study presents a two-stage object detection system specifically tailored for low-light conditions. In the initial stage, supervised deep learning image enhancement techniques are utilized to improve image quality and enhance features. The second stage employs a computer vision algorithm for object detection. Three image enhancement algorithms—ZeroDCE++, Gladnet, and two-branch exposure-fusion network for low-light image enhancement (TBEFN)—were assessed in the first stage to enhance image quality. YOLOv7 was utilized in the object detection phase. The ExDark dataset, recognized for its extensive collection of low-light images, served as the basis for training and evaluation. No-reference image quality evaluators were applied to measure improvements in image quality, while object detection performance was assessed using metrics such as recall and mean average precision (mAP). The results indicated that the two-stage system incorporating TBEFN significantly improved detection performance, achieving a mAP of 0.574, compared to 0.49 for YOLOv7 without the enhancement stage. Furthermore, this study investigated the relationship between object detection performance and image quality evaluation metrics, revealing that the image quality evaluator NIQE exhibited a strong correlation with mAP for object detection. This correlation aids in identifying the features that influence computer vision performance, thereby facilitating its enhancement. |
| format | Article |
| id | doaj-art-e8b8b1b8b17e4e1bbc3a5694f816e30d |
| institution | OA Journals |
| issn | 2376-5992 |
| language | English |
| publishDate | 2025-04-01 |
| publisher | PeerJ Inc. |
| record_format | Article |
| series | PeerJ Computer Science |
| spelling | doaj-art-e8b8b1b8b17e4e1bbc3a5694f816e30d2025-08-20T02:08:07ZengPeerJ Inc.PeerJ Computer Science2376-59922025-04-0111e279910.7717/peerj-cs.2799Two-stage object detection in low-light environments using deep learning image enhancementGhaith Al-refai0Hisham Elmoaqet1Abdullah Al-Refai2Ahmad Alzu’bi3Tawfik Al-Hadhrami4Abedalrhman Alkhateeb5Department of Mechatronics Engineering, German Jordanian University, Amman, JordanDepartment of Mechatronics Engineering, German Jordanian University, Amman, JordanSoftware Engineering Department, King Hussein School of Computing Science, Princess Sumaya University College for Technology, Amman, JordanDepartment of Computer Science, Faculty of Computer and Information Technology, Jordan University of Science and Technology, Irbid, JordanDepartment of Computer Science, School of Science and Technology, Nottingham Trent University, Nottingham, United KingdomDepartment of Computer Science, Lakehead University, Thunder Bay, CanadaThis study presents a two-stage object detection system specifically tailored for low-light conditions. In the initial stage, supervised deep learning image enhancement techniques are utilized to improve image quality and enhance features. The second stage employs a computer vision algorithm for object detection. Three image enhancement algorithms—ZeroDCE++, Gladnet, and two-branch exposure-fusion network for low-light image enhancement (TBEFN)—were assessed in the first stage to enhance image quality. YOLOv7 was utilized in the object detection phase. The ExDark dataset, recognized for its extensive collection of low-light images, served as the basis for training and evaluation. No-reference image quality evaluators were applied to measure improvements in image quality, while object detection performance was assessed using metrics such as recall and mean average precision (mAP). The results indicated that the two-stage system incorporating TBEFN significantly improved detection performance, achieving a mAP of 0.574, compared to 0.49 for YOLOv7 without the enhancement stage. Furthermore, this study investigated the relationship between object detection performance and image quality evaluation metrics, revealing that the image quality evaluator NIQE exhibited a strong correlation with mAP for object detection. This correlation aids in identifying the features that influence computer vision performance, thereby facilitating its enhancement.https://peerj.com/articles/cs-2799.pdfLow-light visionComputer visionCNNImage enhancementYOLOTwo-stage object detection |
| spellingShingle | Ghaith Al-refai Hisham Elmoaqet Abdullah Al-Refai Ahmad Alzu’bi Tawfik Al-Hadhrami Abedalrhman Alkhateeb Two-stage object detection in low-light environments using deep learning image enhancement PeerJ Computer Science Low-light vision Computer vision CNN Image enhancement YOLO Two-stage object detection |
| title | Two-stage object detection in low-light environments using deep learning image enhancement |
| title_full | Two-stage object detection in low-light environments using deep learning image enhancement |
| title_fullStr | Two-stage object detection in low-light environments using deep learning image enhancement |
| title_full_unstemmed | Two-stage object detection in low-light environments using deep learning image enhancement |
| title_short | Two-stage object detection in low-light environments using deep learning image enhancement |
| title_sort | two stage object detection in low light environments using deep learning image enhancement |
| topic | Low-light vision Computer vision CNN Image enhancement YOLO Two-stage object detection |
| url | https://peerj.com/articles/cs-2799.pdf |
| work_keys_str_mv | AT ghaithalrefai twostageobjectdetectioninlowlightenvironmentsusingdeeplearningimageenhancement AT hishamelmoaqet twostageobjectdetectioninlowlightenvironmentsusingdeeplearningimageenhancement AT abdullahalrefai twostageobjectdetectioninlowlightenvironmentsusingdeeplearningimageenhancement AT ahmadalzubi twostageobjectdetectioninlowlightenvironmentsusingdeeplearningimageenhancement AT tawfikalhadhrami twostageobjectdetectioninlowlightenvironmentsusingdeeplearningimageenhancement AT abedalrhmanalkhateeb twostageobjectdetectioninlowlightenvironmentsusingdeeplearningimageenhancement |